This research examined: (1) the effectiveness of learning through a goal-free problem strategy compared with learning through a goal-given problem strategy with regard to transfer ability, reasoning, or cognitive load; (2) the effectiveness of learning collaboratively compared with learning individually with regard to transfer ability, reasoning, or cognitive load; and (3) interaction between learning strategy (goal-free or goal-given) and setting (collaborative or Individual) in terms of transfer ability, reasoning, or cognitive load. The experiment was conducted using a 2 (setting: collaborative vs. Individual) x 2 (strategy: goal-free vs. goal-given) factorial design; posttest-only control-group. The results of the data analysis supported hypothesis that the students who learned by goal-free problems strategy led to be better in terms of near and far transfer scores than students who learned by the goal-given strategy. The goal-free problems strategy was also proved to be more effective in terms of students' reasoning in near and far transfer tests rather than the goal-given problems strategy. Collaborative learning more effectively minimizes cognitive load during acquisition and near transfer test phase than students learn individually, also effectively facilitating students in terms of the near transfer only than students in individual learning. There was a significant interaction effect between the learning setting factor and strategy factor in terms of near transfer ability and reasoning in the near transfer test. Simple effect test showed that learning collaboratively by goal-given strategy was significantly more effective in terms of near transfer ability and students’ reasoning in the near transfer test, and learning individually by goal-free strategy was significantly more effective in terms of near transfer ability and students’ reasoning in near transfer test.